One of the major sources of bias in force rebalanced accelerometers is due to the spatial separation between the pickoff electrical null and the proofmass flexure suspension mechanical null. In closed loop operation the force rebalance servo maintains the accelerometer proofmass at the pickoff null. The force required to hold the proofmass at the pickoff null against the combined forces of the mechanical and electrostatic springs of the deflected flexure suspension is seen as a bias error in the output of the accelerometer. In the current state-of-the-art, the accelerometer bias is measured and modeled over temperature during final calibration. The model is subsequently used to correct the measured acceleration for bias over temperature. The residual to the thermal model is an error that represents one of the limits to the performance capability of the inertial system. This residual error may be due to hysteresis or instability in the mechanical, electrical and thermal characteristics of the accelerometer. It is therefore desirable to minimize accelerometer bias and bias variation to improve inertial system performance, increase yield and reduce cost.